class: center, middle, inverse, title-slide

.title[
# 
]
.author[
### 
]

---

class: center, middle, animated, bounceInDown

#### Theory lessons

| Marta Coronado Zamora | Jose F. Sánchez | 
|:-:|:-:|
| <a href="mailto:Marta.coronado@uab.cat"> marta.coronado@uab.cat</a> | <a href="mailto:JoseFrancisco.Sanchez@uab.cat">&nbsp; josefrancisco.sanchez@uab.cat</a> | 
| <a href="https://bsky.app/profile/geneticament.bsky.social">&nbsp; @geneticament</a> | <a href="https://twitter.com/JFSanchezBioinf">&nbsp; @JFSanchezBioinf</a> |
| <a href="https://portalrecerca.uab.cat/es/organisations/grup-de-gen%C3%B2mica-bioinform%C3%A0tica-i-biologia-evolutiva-gbbe/">&nbsp; Universitat Autònoma de Barcelona </a> | <a href="http://www.germanstrias.org/technology-services/genomica-bioinformatica/"> Germans Trias i Pujol Research Institute (IGTP)</a> |

#### Practical lessons

| Miriam Merenciano |
|:-:|
| <a href="mailto:miriam.merenciano@uab.cat"> miriam.merenciano@uab.cat </a> | 
| <a href="https://portalrecerca.uab.cat/es/organisations/grup-de-gen%C3%B2mica-bioinform%C3%A0tica-i-biologia-evolutiva-gbbe/">&nbsp; Universitat Autònoma de Barcelona </a> |

---
layout: true
class: animated, fadeIn

---

# Theory session dynamics

__Content__

- Theory 
- Examples and short exercises in class () - complete and submit to [Atenea](https://atenea.upc.edu/course/view.php?id=105474)

_Bring your laptop to theory sessions!_

**Interactive documents** 
`R code` can be executed within RStudio!

``` r
value <- 2
value + 3
```

```
## [1] 5
```

---
layout: false
class: left, bottom, inverse, animated, bounceInDown

# Get started!
## Graphics with `ggplot2`

---
layout: true
class: animated, fadeIn

---
# The grammar of graphics

.pull-left[
#### Original grammar
 Wilkinson, Leland. The grammar of graphics. Springer Science & Business Media, 2006.
]

.pull-right[
<center>
![](data:image/png;base64,#img/books.png)
]

#### Adapted to `R` in the `ggplot2` package

Hadley Wickham. ggplot2: elegant graphics for data analysis. Springer, 2009.
 
"The grammar tells us that a __statistical graphic is a mapping from data to aesthetic attributes__ (colour, shape, size) __of geometric objects__ (points, lines, bars). The plot may also contain __statistical transformations__ of the data and is drawn on a specific __coordinate system__. __Facetting__ can be used to generate the same plot for different subsets of the dataset. It is the combination of these independent components that make up a graphic."

---

## Syntax

In `ggplot2` there are different components we can add to a plot:

.pull-left[
```r
ggplot(data = <DATA>,
 mapping = aes(<MAPPINGS>)) +

<GEOM_FUNCTION>(stat = <STAT>,
 position = <POSITION>) +
 
 <SCALE_FUNCTION>() +
 
 <COORDINATE_FUNCTION>() +
 
 <FACET_FUNCTION>() +
 
 <THEME_FUNCTION>
```

]

<img src="img/ggplot2layers2019_8_10D11_28_23.png" width="52%" style="position: absolute; right: 50px; top: 25%;">
<center>

Grammar of Graphics: A layered approach to elegant visuals

---

## Components

The layered grammar defines a plot as a combination of:

- __Layers__
    + Data
    + Aesthetic mapping
    + Geometric objects
    + Statistical transformation
    + Position adjustment
- __Scales__
- __Coordinate system__
- __Faceting specification__
- __Theme__ (not in the original grammar)

<center>

Grammar of Graphics: A layered approach to elegant visuals

---

### Layers

**Layers** are responsible for creating the objects that we perceive on the plot.

- __Data__
- __Aesthetic mapping__
- __Geometric objects__ 
- __Statistical transformation__
- __Position adjustment__

#### Syntax:

``` r
ggplot(data = data,  mapping = aes(x = var1, y = var2, colour = var3)) +
  layer(geom = "point", stat = "identity", position = "identity")
```

---

### Layers: Data

The data layer specifies the data being plotted. Must be an **R data frame** object.

``` r
head(iris, 5)
```

```
##   Sepal.Length Sepal.Width Petal.Length Petal.Width Species
## 1          5.1         3.5          1.4         0.2  setosa
## 2          4.9         3.0          1.4         0.2  setosa
## 3          4.7         3.2          1.3         0.2  setosa
## 4          4.6         3.1          1.5         0.2  setosa
## 5          5.0         3.6          1.4         0.2  setosa
```

``` r
ggplot(data = iris)
```

<img src="data:image/png;base64,#1_files/figure-html/unnamed-chunk-4-1.png" alt="" width="180" style="display: block; margin: auto;" />
We get a blank square because we have not added any other layers yet!

---

### Layers: mapping data onto aesthetics

The `iris` data frame has different columns: **aesthetics**.

The aesthetic layer, or aes for short, specifies how we want to map our data onto the scales of the plot (such as the `x` and `y` coordinates).

In `ggplot2` the aesthetic layer is specified using the `aes()` function.

``` r
ggplot(data = iris, mapping = aes(x = Sepal.Width, y = Sepal.Length))
```

The graph shows variable and scales of `Sepal.Length` mapped onto the `x`-axis and `Sepal.Width` on the `y`-axis. What are we missing?

A **geom**!

---
### Layers: mapping data onto aesthetics

The **geom** is the geometric object to use display the data.

``` r
ggplot(data = iris, mapping = aes(x = Sepal.Width, y = Sepal.Length)) +
  geom_point()
```

Now we have a scatterplot of the relationship between `Sepal.Length` and `Sepal.Width`.

---

### Layers: mapping data onto aesthetics
#### Main aesthetics types:

#### **Exercise** | Which aesthetic attributes can a continuous variable be mapped to? And a discrete variable?

---
### Layers: mapping data onto aesthetics

Data and mapping can be defined in the initial `ggplot()` call or in the layer (`geom_` or `stat_`)

``` r
# Default data and mapping used by the layer
ggplot(data = iris, mapping = aes(x = Sepal.Width, y = Sepal.Length)) +
  geom_point()
# No default defined, data and mapping in layer
ggplot() +
  geom_point(data = iris, mapping = aes(x = Sepal.Width, y = Sepal.Length))
```

#### **Exercise** | With the code that generate the previous figure, experiment with the colour, size, transparency (<code>alpha</code>) and shape aesthetics.

<div style="background-color:#F0F0F0">
&emsp;
Answer:

&emsp;
</div>

---

### Layers: mapping data onto aesthetics

All variables used in the plot should be in the data.

``` r
# Correct
ggplot(iris, aes(x = Sepal.Width, y = Sepal.Length)) + geom_point()

# Not recommended
ggplot(iris, aes(x = Sepal.Width, y = iris$Sepal.Length)) + geom_point()
SL_var <- iris$Sepal.Length
ggplot(iris, aes(x = Sepal.Width, y = SL_var)) + geom_point()
```

---
### Layers: mapping data onto aesthetics

#### **Exercise** | Simplify the following code.

``` r
ggplot(mpg) +
  geom_point(aes(mpg$displ, mpg$hwy))

ggplot() +
  geom_point(mapping = aes(y = hwy, x = cty), data = mpg) +
  geom_smooth(data = mpg, mapping = aes(cty, hwy))

ggplot(diamonds, aes(carat, price)) +
  geom_point(aes(log(brainwt), log(bodywt)), data = msleep)
```

---

### Layers: mapping data onto aesthetics

#### Mapping (variable) vs. setting (constant)

``` r
p <- ggplot(iris, aes(Sepal.Width, Sepal.Length))

p + geom_point(aes(colour = Species)) # Map Species to colour
p + geom_point(aes(colour = "darkblue")) # Map "darkblue" to colour
p + geom_point(colour = "darkblue") # Set colour to "darkblue"
```

---

### Layers: mapping data onto aesthetics

__Group__ aesthetic: by default it is the combination of discrete variables (except position).

#### Multiple groups, one aesthetic:

``` r
ggplot(nlme::Oxboys, aes(age, height)) +
  geom_line()

ggplot(nlme::Oxboys, aes(age, height, group = Subject)) + # or colour = Subject
  geom_line()
```

---

### Layers: mapping data onto aesthetics

__Group__ aesthetic: by default it is the combination of discrete variables (except position).

#### Different groups on different layers:

``` r
ggplot(nlme::Oxboys, aes(age, height, group = Subject)) +
  geom_line() + geom_smooth(aes(group = Subject))

ggplot(nlme::Oxboys, aes(age, height, group = Subject)) +
  geom_line() + geom_smooth(aes(group = 1)) # o group = NULL
```

---
### Layers: mapping data onto aesthetics

#### **Exercise** | Are the observations grouped in the following plot? If so, what variables are used for the grouping?

---
### Layers: mapping data onto aesthetics

- Aesthetic attributes:
    + __Position__: x, y, xmin, xmax, ymin, ymax, xend, yend
    + __Colour__: colour, fill, alpha
    + __Differentiation__: shape, size, linetype
    + __Grouping__: group

- Each geom understands a different set of aesthetics
- Each geom requires some aesthetics (`?geom_*`)

---
### Layers: geometric objects

#### **Exercise** | Which geoms are collective? (multiple observations share a geometrical object)

---

### Layers: geometric objects

- Full list of geom_* functions:
    + [ggplot2 reference](https://ggplot2.tidyverse.org/reference/index.html#section-layer-geoms)
    + [ggplot2 cheat sheet](https://raw.githubusercontent.com/rstudio/cheatsheets/main/data-visualization.pdf), sorted by type

---

### Layers: statistical transformation

The statistics layer allows you plot statistical values calculated from the data. Transforms the data, typically by summarising  it  in  some  manner.

A stat takes a dataset as input and returns a dataset as output, and so a stat can add new variables to the original dataset.

<center>
dataset statistical transformation dataset
</center>

``` r
ggplot(diamonds, aes(carat)) + geom_histogram(aes(y=after_stat(count)), binwidth = 0.1)
ggplot(diamonds, aes(carat)) + geom_histogram(aes(y=after_stat(density)), binwidth = 0.1)
```

New variables are accessed with `after_stat(var)` notation or `stat()` (v3.4.0+)

---

### Layers: statistical transformation

``` r
p <- ggplot(mtcars, aes(x = factor(cyl)))

p + geom_bar() + labs(title = "count") # uses stat = "bin" by default
p + geom_bar(aes(y = after_stat(count)/sum(after_stat(count)))) +
  labs(title = "count/sum(count)")
# p + geom_bar(aes(y = stat(count/sum(count))))
p + geom_bar(aes(y = after_stat(prop), group = 1)) + labs(title = "prop")
```

Note that we need to specify `group = 1` (or `group = NULL`), otherwise the proportion are calculated within each group and therefore all proportions would be 1.

---

### Layers: statistical transformation

- Full list of stat_* functions:
    + `ggplot2` reference: [geoms](https://ggplot2.tidyverse.org/reference/index.html#section-layer-geoms) and [stats](https://ggplot2.tidyverse.org/reference/index.html#section-layer-stats)
    + [`ggplot2` cheat sheet](https://raw.githubusercontent.com/rstudio/cheatsheets/main/data-visualization.pdf)

- New variables from statistical transformations:
    + `stat_count`: count, prop
    + `stat_bin` or `stat_bin2d`: count, density
    + `stat_boxplot`: width, ymin, ymax, middle
    + `stat_smooth`: y, ymin, ymax, se
    + `stat_summary`: value
    
---

### Layers: position adjustment

Position adjustments apply minor tweaks to the position of elements within a layer.

``` r
p <- ggplot(mtcars, aes(x = factor(cyl), y = mpg, colour = factor(cyl)))

p + geom_point() + labs(title = "identity") # default point position
p + geom_point(position = "jitter") + labs(title = "jitter")
p + geom_point(position = position_nudge(x = 0.3)) + labs(title = "nudge")
```

---

### Layers: position adjustment

Position adjustments apply minor tweaks to the position of elements within a layer.

``` r
b <- ggplot(mtcars, aes(x = factor(cyl), fill = factor(gear)))

b + geom_bar() + labs(title="stack") # default bar position
b + geom_bar(position = "fill") + labs(title="fill")
b + geom_bar(position = "dodge") + labs(title="dodge")
```

---

### Layers: data + aes + geom + stat + position

|fun         |aes                                 |geom   |stat     |position |
|:-----------|:-----------------------------------|:------|:--------|:--------|
|geom_point  |x, y, alpha, colour, size, shape... |point  |identity |identity |
|geom_line   |x, y, linetype...                   |line   |identity |identity |
|geom_ribbon |x, ymin, ymax...                    |ribbon |identity |identity |
|geom_col    |x, y, ...                           |col    |identity |stack    |
|geom_text   |x, y, label, angle, fontface...     |text   |identity |identity |
|stat_ecdf   |x, colour, linetype                 |step   |ecdf     |identity |

---

###  Layers: type of layers and annotations

We can add additional **metadata**, context and annotations, that help to give meaning to the raw data or highlight important features.

``` r
ggplot(iris, aes(x = Species, y = Petal.Width, colour = Species)) + 
  # Show data
  geom_jitter(show.legend = FALSE) +
  
  # Summarize
  geom_boxplot(fill = NA, show.legend = FALSE) +
  
  # Annotate
  geom_hline(yintercept = mean(iris$Petal.Width), alpha = 0.5)
```

---

### Layers: type of layers and annotations

- Common `geom_*` functions with metadata:
    + `geom_text()`, `geom_rect()`, `geom_point()`
    + `geom_line()`, `geom_path()`, `geom_segment()`

.pull-left[

``` r
p <- ggplot(mtcars, aes(hp, mpg)) + geom_point()
p + 
 # Add red cercles
 geom_point(data = mtcars[mtcars$hp > 300 | mtcars$mpg > 30,],
 colour = "red", shape = 1, size = 7) +
 # Draw an arrow
 geom_segment(data = mtcars[mtcars$hp > 100 & mtcars$mpg > 30,],
 aes(xend = hp + 30, yend = mpg + 3),
 arrow = arrow(ends = "first", type = "closed", angle = 15, length = unit(0.5,"cm")),
 colour = "gray50") + 
 # Add text
 geom_text(data = mtcars[mtcars$hp > 100 & mtcars$mpg > 30,],
 aes(label = "Economical car"), position = position_nudge(x = 30, y = 3), hjust = 0)
```
]

.pull-right[ 
<img src="data:image/png;base64,#1_files/figure-html/unnamed-chunk-25-1.png" alt="" width="432" style="display: block; margin: auto;" />
]

---

### Layers: type of layers and annotations

- Special layers that act as normal geoms if parameters are within `aes()` or as annotations if outside.
    + `geom_vline()`, `geom_hline()`, `geom_abline()`
- Special layers that don't inherit global settings
    + `annotate()`, `annotation_custom()`

.pull-left[

``` r
# Represent mpg vs. hp and annotate a subset of points with a rectangle
p_overview <- ggplot(mtcars, aes(hp, mpg)) +
 annotate(geom = "rect", xmin = 50, ymin = 25, xmax = 100, ymax = 35, fill = "red", alpha = 0.3 )+
 geom_point() +
 labs(x = NULL, y = NULL)
# Represent the subset of points
p_subset <- ggplot(mtcars[mtcars$hp <100 & mtcars$mpg >25,], aes(hp, mpg)) +
 geom_point()
# Add an inset with the full data
p_subset +
 annotation_custom(ggplotGrob(p_overview), xmin = 75, ymin = 30)
```
]

.pull-right[ 
<img src="data:image/png;base64,#1_files/figure-html/unnamed-chunk-27-1.png" alt="" width="432" style="display: block; margin: auto;" />
]

---

### Layers: type of layers and annotations

#### **Exercise** | Exercise: what type of annotation is the most appropriate for the following cases:

A. To highlight some points

B. To add the expected linear trend

C. To write some text

---

### Scales

Scales control the mapping from data to aesthetics. They take your data and turn it into something that you can perceive visually: e.g., size, colour, position or shape.

- Types of scales
 + Axis: `x` and `y` position
 + Legend: the other aesthetics (fill, alpha, shape...)
- Default scales
 + `scale_<aes>_continuous()`: map continious values to visual values
 + `scale_<aes>_discrete()`: map discrete values to visual values
- Manual scales for discrete variables
 + `scale_<aes>_manual(values = c(1, 2, 4))`: map pdiscrete values to manualy chosen visual values
 
---

### Scales

In all scales you can define: `name`, `breaks` and `labels`.

``` r
ggplot(mtcars, aes(x = mpg, y = wt, shape = factor(gear))) +
  geom_point() +
  scale_y_continuous(name = "Weight", breaks = c(2.5,5)) +
  scale_x_continuous(name = quote(frac(miles, gallon))) +
  scale_shape_discrete(name = "Type", labels = c("III","IV","V"))
```

---

### Scales: legends

- Can be turned on/off
- Need the same name, breaks and labels to be merged

``` r
p <- ggplot(mtcars, aes(x = mpg, y = wt, shape = factor(gear), colour = factor(gear)))
p + geom_point()
p + geom_point(show.legend = FALSE)
p + geom_point() +
 scale_shape_discrete(name = "Type", labels = c("III","IV","V")) +
 scale_colour_discrete(name = NULL)
```

---

### Scales: continuous axes

``` r
p <- ggplot(mtcars, aes(hp, mpg)) + 
 geom_point()
p + labs(title = "untransformed")
p + scale_x_continuous(trans = "log10") + labs(title = "log(x)")
p + scale_y_reverse() + labs(title = "rev(x)")
```

---

### Scales: colour scales for discrete variables

---

### Scales: colour scales for continuous  variables

---

### Coordinate system

The coordinate component allows you to adjust the `x` and `y` coordinates.

<img src="data:image/png;base64,#1_files/figure-html/unnamed-chunk-34-1.png" alt="" width="288" /><img src="data:image/png;base64,#1_files/figure-html/unnamed-chunk-34-2.png" alt="" width="288" /><img src="data:image/png;base64,#1_files/figure-html/unnamed-chunk-34-3.png" alt="" width="288" /><img src="data:image/png;base64,#1_files/figure-html/unnamed-chunk-34-4.png" alt="" width="288" /><img src="data:image/png;base64,#1_files/figure-html/unnamed-chunk-34-5.png" alt="" width="288" /><img src="data:image/png;base64,#1_files/figure-html/unnamed-chunk-34-6.png" alt="" width="288" />
---

### Coordinate system
####Zoom

``` r
p <- ggplot(data = data.frame(x = c(1,2.5,3,3.5,5,3.5,3,2.5,1),
 y = c(3,3.5,5,3.5,3,2.5,1,2.5,3)),
 aes(x, y)) +
 geom_polygon()
p + labs(title = "Full")
p + coord_cartesian(xlim = c(2,4)) + labs(title = "Image clipped")
```

---
### Coordinate system
#### Maps

``` r
# install.packages("maps"); library(maps)
if(require("maps")){
 worldmap <- map_data(map = "world")
 ggplot(worldmap, aes(x = long, y = lat, group = group)) +
 geom_polygon(fill = "gray20", colour = "gray92") +
 coord_quickmap()
}
```

- `coord_map` can use projections from package `mapproj` (needs to be installed)

---

### Facets

Faceting is a mechanism for automatically laying out multiple plots on a page. It splits the data into subsets, and then plots each subset into a different panel on the page. Such plots are often called small multiples.

``` r
p <- ggplot(mtcars, aes(x = wt, y = mpg, colour = factor(cyl),
 shape = factor(am))) +
 geom_point(show.legend = FALSE) + labs(x = NULL, y = NULL)
p 
p + facet_wrap(~factor(cyl)) # 1d ribbon of panels that is wrapped into 2d
p + facet_grid(factor(am) ~ factor(cyl)) # 2d grid of panels defined by variables which form the rows and columns
```

---

### Facets

`facet_grid` have two additional arguments: `scales` and `space`.

``` r
msleep$name <- factor(msleep$name, levels = msleep$name[order(msleep$sleep_total)])
p <- ggplot(msleep[msleep$vore %in% c("carni","insecti"), ], 
 aes(x = sleep_total, y = name)) +
 geom_point() + scale_x_log10()
p + facet_grid(vore~.)
p + facet_grid(vore~., scales = "free", space = "free")
```

---

### Themes

- Complete themes `theme_*()`
- Theme elements `theme()`

- Full list of theme elements: `?theme`

---

## Components

The layered grammar defines a plot as a combination of:

<center>

Grammar of Graphics: A layered approach to elegant visuals

---

## Components

#### **Exercise** | Update the plot to match the requirements below

``` r
p <- ggplot(msleep, aes(x = vore, y = sleep_total, colour = vore,
 fill = vore)) +
 geom_violin(alpha = 0.2, colour = NA) +
 geom_jitter(width = 0.3)
```

- Legend on top of the plot
- White background
- No ticks in y axis
- Axis labels rotated to an angle of 45 degrees

Note: check `?theme` options

---

## Components

#### **Exercise** | Update the plot to match the requirements below

---
layout: false
class: inverse, center, middle, animated, bounceInDown

### Upload `1.Rmd` with the completed exercises (text included) to [Atenea](https://atenea.upc.edu/course/view.php?id=105474)